**2.2 Analytical methods**

The content of H2, CO2 was measured with gas chromatography (Varian GC-3800 equipped with Carboplot P7 capillary column and TCD). The loss of organic substances was monitored with COD measurement (dichromate method) after centrifugation of biomass (Standard methods, 1995). The biomass content was established spectrophotometrically measuring optical density at 660 nm (DU640 UV-VIS spectrophotometer from Beckmann). Cell dry weight was determined using gravimetric method. Six samples from the same kinetic measurement points at respective time intervals were mixed together, 10 ml of cell suspension was centrifuged at 12000 g for 12 min, the pellet was washed twice with deionized water and dried at 80oC for 4 h. Elemental analysis of the foot wastes (C,H,N,O) was performed in triplicate using an elemental analyser (Vario EL III Elementary). Concentration of Fe, Ca, Mg in purified wastes was measured by ICP OES spectroscopy. The value of pH was measured with glass electrode ERH-11. The intensity of luminance was measured at the external wall of the bottle with a luxometer Lx204 made by Slandi, Poland and a pyranometer CMP3 by Kipp & Zonen (Waligórska, 2006). The light conversion efficiency (η) was calculated based on the following formula (Koku, 2002):

$$
\eta(\%) = \frac{\\$3.61 \cdot \rho \cdot V}{I \cdot A \cdot t} \tag{1}
$$

where "V" is the volume of produced H2 in liters, "ρ" is the density of the produced hydrogen gas in g/l, "I" is the light intensity in W/m2, "A" is the irradiated area in m2 and "t" is the duration of hydrogen production in hours.

Substrate efficiency Ysub (l /l waste) was calculated as final hydrogen concentration per l of waste:

$$Y\_{sub} = \frac{H\_{\text{max}}}{V\_{waste}} \tag{2}$$

where Hmax is a final hydrogen concentration in l, Vwaste is waste concentration in l.

Specific efficiency Ysp (l H2 /g COD) was calculated based on following equation:

$$Y\_{sp} = \frac{H\_{\text{max}}}{\text{COD}\_{\text{loss}}} \tag{3}$$

The modified Gompertz (Eq. 4) was applied for calculations of cumulative amounts of hydrogen and carbon dioxide (Mu, 2007, Nath, 2008, Chen, 2006):

$$H = H\_{\text{max}} \exp\left\{-\exp\left[\frac{R\_{\text{max},H\_2}e}{H\_{\text{max}}}(\mathcal{A}-t)+1\right]\right\} \tag{4}$$

where: H - cumulative hydrogen (l/lmedium), Hmax – maximum cumulative hydrogen (l/lmedium), Rmax, H2 –maximum rate of hydrogen production (l/l/h), t – fermentation time (h), λ – lag time (h), e – exp = 2.718.

#### **3. Results and discussion**

#### **3.1 Pretreatment of wastes**

252 Biogas

simultaneous degradation of these laborious wastes seems to be a very environmentally friendly solution. The US Department of Energy Hydrogen Program in United States estimates that contribution of hydrogen to total energy market will be 8-10% by 2025 (National Hydrogen Energy Roapmap, 2002). It is predicted that hydrogen will become the main carrier of energy in the near future due to environmental and universal applications reasons. It is clean, highly energetic energy carrier (142.35 kJ/g), with almost tripled gravimetric energy density compared to ordinary hydrocarbons. Although the described method is relatively simple and cheap it still requires optimization due to the obtained

Photoheterotrophic bacteria *Rhodobacter sphaeroides* O.U. 001 ATTC 4919 (Fig.1) were cultivated on Van Niel's medium containing: K2HPO4 (1.0 g/l), MgSO4 (0.5 g/l), yeast extract (10g/l) and tap water filled up to 1 l and then activated according to the procedure already described (Waligórska, 2006). For hydrogen generation a modified Biebl and Pfennig medium (Biebl, 1981) was applied. This standard medium contained: KH2PO4 (0.5 g/l); MgSO4\*7H2O (0.2 g/l); NaCl (0.4 g/l); CaCl2\*2H2O (0.05 g/l), L-malic (2.0 g/l); sodium glutamate (0.36 g/l), ferric tartrate (0.005 g/l); yeast extract (0.17 g/l) and microelements: ZnCl2 (0.07 g/l); MnCl2\*4H2O (0.1 g/l); H3BO3 (0.06 g/l); CoCl2\*6H2O (0.2 g/l); CuCl2\*2H2O

The untreated food waste were initially filtered through cotton wool, next sterilized at 120oC

Wastes with different COD values (46 g O2/l for dairy wastes, 220 and 27 g O2/l for brewery wastes) after pretreatment were introduced to the medium, which did not contain L-malic acid. The medium was inoculated with bacteria 30% v/v (0.36 g dry wt/l). The process was performed in small vials (25 ml) made from sodium glass and filled with 12.5 ml of inoculated medium. Tightly closed vials were carefully deaerated with argon before starting the illumination. All experiments were carried out at 28 ± 2oC and pH after sterilization and inoculation varied between 7.0 and 7.2. The mercury-tungsten lamp (Ultra-Vitalux –300W from Osram) was applied in all experiments. The intensity of light during hydrogen generation was 9 klx (116 W/m2). The vials with Biebl and Pfenniga standard medium was

The content of H2, CO2 was measured with gas chromatography (Varian GC-3800 equipped with Carboplot P7 capillary column and TCD). The loss of organic substances was monitored with COD measurement (dichromate method) after centrifugation of biomass (Standard methods, 1995). The biomass content was established spectrophotometrically measuring optical density at 660 nm (DU640 UV-VIS spectrophotometer from Beckmann). Cell dry weight was determined using gravimetric method. Six samples from the same kinetic measurement points at respective time intervals were mixed together, 10 ml of cell suspension was centrifuged at 12000 g for 12 min, the pellet was washed twice with

(0.02 g/l); NiCl2\*6H2O (0.02 g/l); Na2MoO4\*2H2O (0.04 g/l); HCl 25% (1ml/1).

by autoclaving for 20 min and re-filtered applying paper filter.

unsatisfied yields.

**2. Materials and methods** 

used as reference (Biebl, 1981).

**2.2 Analytical methods** 

**2.1 Inoculum, medium and procedures** 

The wastes applied in this series of experiments required high temperature pretreatment (120oC for 20 min), which had significantly increased the efficiency of hydrogen production by removing from the crude waste microorganisms realizing competitive fermentation. The crude wastes were acidic (dairy waste pH 4.2, brewery waste pH 4.7) and contained high concentration of NH4+ (40 mg/l dairy waste and 96 mg/l brewery waste), which can significantly reduce hydrogen production (Waligórska, 2009). High concentration of

not observed.

3,0

hydrogen volume [l/l medium]

0,0

(Seifert, 2010)

**3.2 Light intensity effect** 

0,5

1,0

1,5

2,0

2,5

Photofermentative Hydrogen Generation in Presence of Waste Water from Food Industry 255

Many food waste, for example dairy waste, contained significant amounts of whey particles. It was interesting to check whether microorganisms utilize only organic compounds from solution or may be originate from consumption of solid particles as well? The results indicated that at higher concentration of waste the amount of generated hydrogen increased about 40 – 60% when we non-filtered waste (Fig 3). The only exception can be observed In waste with lower concentration 5 % v/v. Here, large differences in hydrogen production are

> 10% sterile waste,30%inoculum 10% unsterile waste,30%inoculum

10% sterile waste,10%inoculum 10% unsterile waste,10%inoculum

time [h] 0 20 40 60 80 100

Fig. 2. Influence of sterilization of brewery wastewaters on kinetics of hydrogen generation.

The whey suspension contains 5 wt.% of lactose, proteins, fats and lactic acid. However, all these components can be an excellent source of organic carbon for *R. sphaeroides* during hydrogen generation (Koku, 2002) due to relatively good solubility. Obeid et al.(2009) used lactic acid as a source of organic carbon in hydrogen generation applying *Rhodobacter capsulatus* and obtained relatively high yield of H2 (5.5 l H2/l) but the acclimatization time was long and lasted 24 h. Sugars, proteins as well as fatty acids were already applied as

For these series of experiments the medium containing 40% v/v of dairy waste and 10% v/v of brewery waste were applied. The media were inoculated with *Rhodobacter sphaeroides*  O.U.001 in concentration 30% v/v 0,36 g dry wt/l. The effect of the light intensity was checked out for 5, 9 and 13 klx. (Fig.4). The highest volumes of hydrogen (3.2 l H2/l medium

substrates in hydrogen photogeneration (Eroglu, 2004; Yokoi, 2002; Zhu, 1999)

N-NH4+ as well as N2 inhibits hydrogen production by nitrogenase. In the absence of nitrogen in the system the nitrogenase catalyses the reduction of protons to molecular hydrogen (Melis, 2006, Yakunin, 1988, Pawlowski, 2003, Dubbs, 2004).

The characteristics of applied wastes is given in Table 1. In order to establish the influence of the wastes pretreatment conditions on the final production of hydrogen a series of experiments with non-treated and sterilized waste were performed. These measurement were performed with solution containing brewery waste at concentration of 10% v/v inoculated with 10% and 30% v/v of inoculums. The results of these experiments are shown on Fig.2. The experiments with "raw", undiluted dairy waste failed.

Fig. 1. *Rhodobacter sphaeroides* ATCC 17032. Micrographs performed with electron microscope with phase contrast (PCM). Tab on left micrograph equals 5 μm (Garrity, 2005).


Table 1. Characteristics of the food wastes.

Application of the sterilized brewery waste with concentration of inoculum 10% v/v resulted in double amount of produced hydrogen. Triplication was observed at higher concentration of inoculums (30% v/v). Many laboratories apply similar pretreatment conditions. Thermal treatment at 95oC for 45 min (Yetis, 2000), filtration or sedimentation (Salih,1989) as well as dilution leads towards removal of fermentation bacteria and solid sediments from medium. Moreover, were applied: illumination with UV radiation, termal treatment at 50oC in presence of 1vol. % of hydrogen peroxide. It was found that only thermal sterilization was successful method.

Many food waste, for example dairy waste, contained significant amounts of whey particles. It was interesting to check whether microorganisms utilize only organic compounds from solution or may be originate from consumption of solid particles as well? The results indicated that at higher concentration of waste the amount of generated hydrogen increased about 40 – 60% when we non-filtered waste (Fig 3). The only exception can be observed In waste with lower concentration 5 % v/v. Here, large differences in hydrogen production are not observed.

Fig. 2. Influence of sterilization of brewery wastewaters on kinetics of hydrogen generation. (Seifert, 2010)

The whey suspension contains 5 wt.% of lactose, proteins, fats and lactic acid. However, all these components can be an excellent source of organic carbon for *R. sphaeroides* during hydrogen generation (Koku, 2002) due to relatively good solubility. Obeid et al.(2009) used lactic acid as a source of organic carbon in hydrogen generation applying *Rhodobacter capsulatus* and obtained relatively high yield of H2 (5.5 l H2/l) but the acclimatization time was long and lasted 24 h. Sugars, proteins as well as fatty acids were already applied as substrates in hydrogen photogeneration (Eroglu, 2004; Yokoi, 2002; Zhu, 1999)

### **3.2 Light intensity effect**

254 Biogas

N-NH4+ as well as N2 inhibits hydrogen production by nitrogenase. In the absence of nitrogen in the system the nitrogenase catalyses the reduction of protons to molecular

The characteristics of applied wastes is given in Table 1. In order to establish the influence of the wastes pretreatment conditions on the final production of hydrogen a series of experiments with non-treated and sterilized waste were performed. These measurement were performed with solution containing brewery waste at concentration of 10% v/v inoculated with 10% and 30% v/v of inoculums. The results of these experiments are shown

hydrogen (Melis, 2006, Yakunin, 1988, Pawlowski, 2003, Dubbs, 2004).

on Fig.2. The experiments with "raw", undiluted dairy waste failed.

Fig. 1. *Rhodobacter sphaeroides* ATCC 17032. Micrographs performed with electron

NH4+ [g/dm3]

Parameters pH COD

Brewery

Brewery

[g/dm3]

Table 1. Characteristics of the food wastes.

thermal sterilization was successful method.

microscope with phase contrast (PCM). Tab on left micrograph equals 5 μm (Garrity, 2005).

waste I 4.7 220 96 0.7 36.7 7 0.05 37.2 1.04 96

waste II 8.5 27 12 0.5 13.3 3 0.01 88.4 0.8 58.6 Dairy waste 4.2 46 40 1.05 35.5 6.3 0.08 1043 0.54 80

Application of the sterilized brewery waste with concentration of inoculum 10% v/v resulted in double amount of produced hydrogen. Triplication was observed at higher concentration of inoculums (30% v/v). Many laboratories apply similar pretreatment conditions. Thermal treatment at 95oC for 45 min (Yetis, 2000), filtration or sedimentation (Salih,1989) as well as dilution leads towards removal of fermentation bacteria and solid sediments from medium. Moreover, were applied: illumination with UV radiation, termal treatment at 50oC in presence of 1vol. % of hydrogen peroxide. It was found that only

C [%]

H [%]

S [%]

Ca [mg/l]

Fe [mg/l]

Mg [mg/l]

N [%]

> For these series of experiments the medium containing 40% v/v of dairy waste and 10% v/v of brewery waste were applied. The media were inoculated with *Rhodobacter sphaeroides*  O.U.001 in concentration 30% v/v 0,36 g dry wt/l. The effect of the light intensity was checked out for 5, 9 and 13 klx. (Fig.4). The highest volumes of hydrogen (3.2 l H2/l medium

Photofermentative Hydrogen Generation in Presence of Waste Water from Food Industry 257

5 klx leads also to high light conversion efficiency. However, in this case the duration of the process significantly increases. For this reason 9 klx seemed to be optimal and has been used

> final time [h]

> > 72 60 60

Table 2. The effect of light intensity on duration of the process, hydrogen production and

In these series of experiments we tested several concentrations of inoculum introduced to the medium: 5-40 % v/v (0.086 g dry wt/l – 0.48 g dry wt/l) for standard medium and 10% and 30% (0.086 g dry wt/l and 0.36 dry wt/l) in case of medium containing wastes (Fig. 5). The optimum inoculum concentration in all cases turned out to be 30% v/v. Data in Table 3 indicate that the second higher concentration produces more hydrogen, shorter lag phase

9klx 13klx

5klx

standard 40% dairy waste 10% brewery waste

η [%]

1.9 2.4 1.7 final time [h]

> 96 60 60

hydrogen [l/l medium]

> 2.0 2.26 2.3

η [%]

1.7 1.7 1.2

hydrogen [l/l medium]

> 1.7 3.2 3.15

for further experiments.

final time [h]

> 106 76 48

hydrogen [l/l medium]

> 2.26 2.3 2.0

light conversion efficiency (η) (30% v/v inoculum).

5klx 9klx

13klx

**3.3 The effect of inoculum concentration** 

η [%]

1.76 1.73 1.33

standard dairy

(30% v/v inoculum, 40% v/v dairy waste, 10% v/v brewery waste)

waste

Fig. 4. The effect of light intensity on hydrogen production in photofermentation process

brewery waste

5klx

9klx 13klx

Light intensity [klx]

> 5 9 13

Hydrogen volume [l/l medium]

0,0

0,5

1,0

1,5

2,0

2,5

3,0

3,5

for dairy wastes, and 2.3 l H2/l medium for brewery wastes) were observed when 9 and 13 klx were applied.

Fig. 3. Influence of filtration of dairy wastewaters on hydrogen generation.

Similar light intensity (8klx) was use by Zhu et al. for tofu wastewaters treated with *Rhodobacter sphaeroides.* Volume of hydrogen obtained for these conditions was 1.5 l H2/l medium (2.8 l H2/l medium when glucose was used) (Zhu, 1999). Nath et al. showed the best results of hydrogen generation when 10 klx was applied (Nath, 2009). Li et al., however, studying the photofermantation of glycerol with *Rhodobacter sphaeroides ZX-5* proved the highest hydrogen production with light intensity not exceeding 5 klx (Li, 2009). Surprisingly, high light intensity was tested by Obeid et al. for photofermentation of lactate medium and *Rhodobacter capsulatus IR3* (up to 50 klx). Highest effectiveness and rate of hydrogen production were obtained when 30-50 klx were used. These tests are essential taking into account that light intensity on sunny day can be higher than 100 klx.

Light intensity seems to be an important factor in hydrogen photogenerating process. On one hand increase at light intensity stimulates hydrogen production and biomass growth, on the other hand too high intensity may cause the reduction of nitrogenase activity or even damage of the cells (Asada, 1999, Uvar, 2005). An important parameter which shows the relationship between light intensity, irradiation area, duration of H2 production and total H2 amount is the light conversion efficiency (η, equation 1). It is the ratio of the total energy of the obtained hydrogen to the total energy input of the photobioreactor by solar radiation (Eroğlu, 2007). In our tests η reached the highest value when 9 klx was applied (2.4 % for dairy waste, 1.7 for brewery waste, Table 2) Results in Table 2 show that illumination with


5 klx leads also to high light conversion efficiency. However, in this case the duration of the process significantly increases. For this reason 9 klx seemed to be optimal and has been used for further experiments.

Table 2. The effect of light intensity on duration of the process, hydrogen production and light conversion efficiency (η) (30% v/v inoculum).
